Man vs. machine: A comparison of human and computer assessment of nonverbal behavior in social anxiety disorder.

Social anxiety disorder (SAD) is a common psychological disorder associated with broad interpersonal impairment. Most previous studies have examined nonverbal behavior in SAD using human coders. However, one recent study utilized a machine-based analysis of nonverbal behavior and dyadic synchrony in SAD (Asher, Kauffmann, & Aderka, 2020). In the present study, we compared human and computer assessments of nonverbal behavior in social anxiety to enhance our knowledge about their commonalities and unique differences in capturing nonverbal behavior in the context of SAD. Specifically, the present study included 152 individuals: 38 individuals diagnosed with SAD and 114 individuals without SAD. Participants formed 76 opposite-sex interaction dyads comprising either two individuals without SAD (n = 39 control dyads) or one individual with SAD and one individual without SAD (n = 37 SAD dyads). All participants underwent a getting-acquainted task and were videotaped during the conversation. Half of the interactions were small talk interactions and half were closeness-generating interactions that required significant self-disclosure. We found that both types of coding were associated with self-reported social anxiety but that machine-based coding was superior in capturing social anxiety in closeness-generating contexts. Implications for research on nonverbal behavior in SAD are discussed.

Emotion regulation during personal goal pursuit: Integration versus suppression of emotions.

OBJECTIVE: Goal pursuit may involve setbacks likely to elicit negative emotions. To continue pursuing the goal, an individual may need to regulate those emotions. In this study, we compared the unique contributions of two emotion regulation styles, integrative emotion regulation (IER) and suppressive emotion regulation (SER), to goal pursuit processes. We tested the hypotheses that IER and SER would be differentially related to goal progress and goal-related effort and goal-related depressed mood would mediate those relations. METHOD: 255 Israeli participants completed five web questionnaires at two-week intervals. We examined the mediation hypothesis using multilevel structural equation modeling. RESULTS: At the within-person level, increases in IER predicted increases in goal progress at a given time point through increases in goal-related effort, while increases in SER predicted decreases in goal progress through increases in goal-related depressed mood. At the between-persons level, participants with higher IER reported more goal progress; this effect was mediated by goal-related effort. Participants with higher SER reported lower goal progress; this effect was mediated by higher goal-related depressed mood. The findings held after controlling for such factors as participants' perceived goal competence, goal stress, sex, and age. CONCLUSIONS: IER promotes goal pursuit, but SER impedes it.